Not applicable
Not applicable
Not applicable
1. Field of the Invention
This invention generally relates to automatic discovery of network topology in multi-node, multi-connection telecommunications networks. In particular, the invention relates to real-time dynamic determination of the physical topology of a network as nodes and/or ports interfacing physical inter-node connections are added to, or deleted from, the network.
2. Description of the Prior Art
Large telecommunications networks are comprised of multiple network elements, each possibly having multiple ports for passing data between the various network elements. A subset of network elements used to transport information may be described as the transmission system. The transmission elements of a telecommunications network transmission system are those elements that interface various transmission links, such as optical fibers, conducting wires or cables, or wireless radio links of various types. The data transported by the transmission system may include: voice, video, digital and analog data in many different formats.
Transmission systems typically include various layers of software. For example, the Open System Interconnection reference model includes seven layers, such as the physical layer. Regardless of the model, the physical layer comprises the various network elements and the associated interconnections. The software drivers for implementing the physical layer direct how bits are placed on and removed from the physical connections between network elements.
System-specific information is transferred between network elements primarily using overhead in the links between ports to communicate among nodes. Overhead may take one or both of two forms. One form is a structure where system control data is defined and formatted to always be present, coexisting with the transmission space allocated to carrying network customer information (i.e. payload), and always allowing a specific amount of customer information to be supported in a given physical path. The various designators in the overhead may change, but the change does not alter the amount of bandwidth or frame capacity dedicated to payload. Examples of standards applied for implementing the physical layer of a communications network are the Synchronous Optical NETwork standard (SONET) and Synchronous Digital Hierarchy (SDH) transmission systems. Another example associated with a data link layer of software is Asynchronous Transfer Mode (ATM) systems.
Another form of overhead is a structure where specifically formatted information is transmitted along with the payload of the network. Examples of this are Resource Management (RM) cells in an ATM system, and Neighbor Information Frames (NIF) in an Fiber Distributed Data Interface (FDDI) network.
One byte of the former type of overhead data, such as defined by SONET, is the Section Trace byte (J0). The Synchronous Digital Hierarchy standard defines the Section Trace byte as a 16 byte message string. As originally intended, the Section Trace byte is repetitively transmitted so that a receiving network element may verify continued connection to the intended transmitting network element. In the case where Network administrators elected not to use the Section Trace byte capability, a 01 Hex is transmitted in the byte.
Regardless of the standards used, transmission systems in large telecommunications networks may change their connectivity characteristics at irregular intervals, such as when new network elements or ports are added to or deleted from the network. Connectivity may also change due to equipment or link failures or maintenance. The current connectivity state of a network is called its topology.
Precise topology information is needed to accomplish many telecommunications network functions. The ability to place a new connection for transferring information from one port to another port, or multiple ports, through a process referred to as circuit provisioning or connection management, is dependent on accurate network topology information. Other dependencies include correlation of network alarms to specific physical locations and restoration of failed connections.
The overall network topology is typically manually entered into a record for use by a management system. If a network element or port is added or removed, the record of the network topology is manually altered to reflect the change. This manual process is subject to human error and requires significant time and resources. Errors result in significant resource expenditures for trouble shooting.
Automatic discovery of the network topology without manual entry of the topology may be provided. These methods rely on transferring topology data between network nodes using data space, such as cells or frames, that might otherwise be used for transmitting customer payload data. One such method is disclosed by Chatwani et al. in U.S. Pat. No. 5,729,685. Data Link layer software, such as Asynchronous Transfer Mode (ATM) protocol software, is used to transmit topology information to the network management system. Link advertisement messages on each of the ports of each ATM switch in the network are transmitted as part of the payload. The messages are received by neighbor switches and forwarded to a topology manager that constructs an overall network topology profile. However, use of the payload bandwidth reduces the amount of bandwidth for transmitting user information. Furthermore, the Data Link layer is removed from the network elements and other hardware.
U.S. Pat. No. 5,481,674 by Mahavadi, describes a method for generating a topology map between devices on an FDDI network. In an FDDI network, a token is passed from controller to controller in a predetermined direction on a path or ring containing all controllers connected to the network. The system determines upstream and downstream neighbors and ports on the FDDI network by performing a mapping based on received Station Information Frame (SIF) responses consecutively sent to elements of the network through exiting connections and ports. The SIF occupies the same information path as the user data.
The above described art reduces the usable traffic capacity (i.e. payload) in a given network link to communicate topology information. The present invention is directed to improvements that allow automatic discovery of network topology without a corresponding reduction in payload bandwidth at the physical layer.
This invention relates to a system and method for determining the topology of a multi-node network such that the method used does not reduce the originally designed information-carrying capacity of the network links, or interrupt existing payload traffic. Overhead data, such as associated with the physical layer, is specifically identified for and is used to transmit unique network and port identifiers from a source node to a destination node connected by a link. The transmission may be continuous.
Each port in a network element has local knowledge of the identity of the corresponding port and network element at the far end of the physical link. A network management system correlates the data in each network element in order to form a topology map for the entire network, allowing the network management system to track changes in links and ports.
Local knowledge in the network element is maintained through the use of object-oriented programming techniques, where the identification of a far-end port is maintained as an attribute associated with the object representing the local port. Changes in this far-end identification attribute causes the object to be updated, resulting in a topology change.
In one aspect described below, a method and system for determining network topology in a communications network is provided. A first network element is connected to a second network element. Data from the first network element is continuously transmitted to the second network element, each transmission of data comprising physical layer overhead data and payload data. A network element identification is provided in the physical layer overhead data without reduction of a bandwidth of the payload data. The network topology is determined from at least the network element identification transmitted to the second network element.
In another aspect described below, a method and system for determining network topology in a communications network comprises: (a) providing first and second network elements, the first network element comprising at least a first port and the second network element comprising at least a second port and a register; (b) connecting the first port to the second port; (c) transmitting a first network element and first port identifier from the first network element to the second network element; (d) storing the first network element and first port identifier in the register; and (e) determining a connection relationship between at least the first and second network elements as a function of the first network element and first port identifier stored in the register.